1 December 2022 in Chaos: An Interdisciplinary Journal of Nonlinear Science
In neuronal network analysis on, for example, synchronization, it has been observed that the influence of interactions between pairwise nodes is essential. This paper further reveals that there exist higher-order interactions among multi-node simplicial complexes. Using a neuronal network of Rulkov maps, the impact of such higher-order interactions on network synchronization is simulated and analyzed. The results show that multi-node interactions can considerably enhance the Rulkov network synchronization, better than pairwise interactions, for involving more and more neurons in the network.
3 January 2023 in Physical Review E
A long-standing expectation is that two repulsively coupled oscillators tend to oscillate in opposite directions. It has been difficult to achieve complete synchrony in coupled identical oscillators with purely repulsive coupling. Here, we introduce a general coupling condition based on the linear matrix of dynamical systems for the emergence of the complete synchronization in pure repulsively coupled oscillators. The proposed coupling profiles (coupling matrices) define a bidirectional cross-coupling link that plays the role of indicator for the onset of complete synchrony between identical oscillators. We illustrate the proposed coupling scheme on several paradigmatic two-coupled chaotic oscillators and validate its effectiveness through the linear stability analysis of the synchronous solution based on the master stability function approach. We further demonstrate that the proposed general condition for the selection of coupling profiles to achieve synchronization even works perfectly for a large ensemble of oscillators.
28 January 2025 in IEEE Transactions on Consumer Electronics
Cone Beam Computed Tomography (CBCT) plays a crucial role in dentistry, providing detailed imaging for diagnosis and treatment planning. However, standard CBCT imaging involves high radiation levels, raising safety concerns and driving the adoption of low-dose imaging, which often compromises image quality. This paper presents a novel denoising pipeline specifically designed to address the complex noise characteristics of low-dose CBCT images, which we have identified as resembling speckle noise. Our approach integrates advanced filtering techniques, innovative noise estimation methods, and brightness correction for 3D image reconstruction, while also leveraging the human visual system’s sensitivity to different frequencies to enhance CBCT visual quality. Experimental results demonstrate that our method outperforms state-of-the-art denoising techniques, including deep learning-based approaches, in achieving superior visual quality. This innovation not only enhances diagnostic precision but also improves patient safety, setting a new benchmark for image quality in dental care.